That's an interesting question and really depends on the following
- How likely do you think a polling error of 2-3 points is.
- How linked are states. IE does a 2 point polling error in Florida mean a 2 point polling error nationally.
- How effective are things like ground game at producing better results than polling.
http://www.nytimes.com/interactive/2...-forecast.html
The upshot site has a great summary of all of the aggregators and modellers forecasts by state.
I think though, that even in a Trump win situation you will see a large initial drop followed by a recovery similar to what happened to the markets around Brexit. So if Trump does win don't sell after the fact as you will be part of a panic sale. If anything if trump wins you should be buying all the way down.
So the odds forcasters range from a 65% chance of Hillary victory to a 99% chance of Hillary victory with the betting markets sitting somewhere at 80%.